Fragment-based discovery and design of bacterial DsbA inhibitors

Abstract

Fragment-Based Drug Discovery (FBDD) is a relatively new approach to drug discovery that initially seeks to detect binding between potential drug targets and small compounds (fragments) having molecular weights between 150 and 300 Da. Fragments are typically small organic molecular “scaffolds” that bind with relatively low absolute affinity but high efficiency. The interactions are detectable only by sensitive biophysical techniques such as Nuclear Magnetic Resonance (NMR) and Surface Plasmon Resonance (SPR) spectroscopy, and X-ray crystallography. Following detection as binders, fragments are elaborated to improve binding affinity while maintaining their ligand efficiency. Theoretically, it is possible by iterative cycles of synthesis and assay to elaborate fragments having millimolar affinity into leads of high nanomolar affinity and ultimately low-nanomolar drugs, while maintaining favourable physicochemical properties. Among the many promising targets for FBDD are bacterial proteins involved in pathogenic virulence. In Gram-negative bacteria, the periplasmic thiol-disulfide oxidoreductase, DsbA, catalyses post-translational disulfide bond formation in a wide range of substrate polypeptides. Functionally, several of these proteins are known to mediate bacterial virulence in host organisms. The existence of a DsbA homologue in the Gram-positive bacterium, Staphylococcus aureus, suggests that it also may owe its pathogenicity to a functional DsbA. Knockout mutations in the prototypic E. coli and other bacterial models have been shown to compromise bacterial virulence. Thus, interruption of DsbA function in other pathogenic bacteria such as Vibrio cholerae could also result in significant impacts on their virulence. The structures of DsbA from S. aureus (SaDsbA) and V. cholerae (VcDsbA) have been solved and the proteins have been functionally characterised through complementation studies in E. coli. As there are no known inhibitors of either SaDsbA or VcDsbA, they represent attractive targets for antibacterial drug discovery by FBDD. The work reported in this thesis involved the application of Fragment-Based Drug Discovery techniques to target validation and to the identification, early elaboration and structural characterisation of potential inhibitors of DsbA from two bacterial sources, Staphylococcus aureus and Vibrio cholerae. Full backbone and side chain resonance assignments of SaDsbA were completed in preparation for subsequent fragment-based binding studies. Ligand- and receptor-based NMR approaches to identify and validate fragment binders of both proteins were employed. Interactions between fragments and SaDsbA were not detected. However, fragment binding with VcDsbA was detected, validated, and studied further by NMR, SPR and X-ray crystallography. The use of small organic solvent molecules as probes of ligand binding sites was explored by NMR, and details of binding between VcDsbA and analogues and derivatives of two of the original fragment hits were investigated further by a range of quantitative and structural NMR techniques. For the structural NMR studies, selective isotopic labelling strategies were employed to generate sparsely proton-labelled VcDsbA to enable the derivation of unambiguous distance restraints. This information was used in data-driven docking calculations to generate structural models of the VcDsbA-fragment complexes. In conclusion, this work identified a small group of aryl fragments and analogues, the most promising of which are based on a trifluoromethylbenzimidazole scaffold, that exhibited affinity for the potential drug target, VcDsbA, representing promising leads for the development of tight-binding inhibitors and hence novel, narrow-spectrum anti-virulence agents.